[PATCH v4] Support ymm rounding control for Intel AVX10.2

[Jan Beulich]

On 20.08.2024 09:21, Jiang, Haochen wrote:
> And I checked the testcases, since I choose from .irp, in those cases, 16-64 bit conversion
> is using the same register type 32-64 bit conversion. (Both xmm -> ymm). That is why it is
> not adding to the cases. If you think they are worth to be added to show the different
> type of conversion, I will add them.

Yes please - at least one example of those (ideally consistent pattern-
wise with how you selected the other subset).

Jan